Headset feedback circuit

Thread Starter

DrJohnPaul

Joined Jan 8, 2021
1
When using a headset with a microphone on PC (actually Mac) applications such as Skype, Zoom, Teams, etc, there is no feedback from the microphone. This means you cannot hear yourself speaking very well - particularly with in-ear or noise-cancelling headsets. This is not an issue with mobile phones (or phones in general) because they all appear to provide feed back. I believe there is a system preference on Windows to provide feedback but not on MacOS. To solve this problem, I set about designing a simple one-stage amplifier that I could use in a small device to intercept the microphone output and feed some of it directly back into the headset. This has turned out to be way more challenging than I expected. There are many types of headset with built-in microphones you can buy, and they all work fine on a laptop. Therefore I presume there must be some agreed standard characteristics.

The first issue is what is the microphone? I assumed it was an electret (condenser) microphone because it is very small and seems to need a power source to provide the charge necessary for any output. 5v produces 20mV signal. But it has resistance (so not a capacitor) and the resistance depends on the polarity of the polarisation, as indeed does the output. This implies there may be some semiconductor stage.

The second issue is that the headphones and microphone use a common connection to the 3.5 mm jack, and this must be +ve with respect to the microphone output for the microphone to work. This in-turn implies I have to use PNP transistor with a +ve ground in the amplifier. This seems a bit odd for mass-produced item.

Admittedly, its been over 30 years since I designed any hardware (and even then it was all digital), but I simply cannot get this thing working. Before I pull apart a headset microphone to try and find out exactly what it going on, does anyone know? Headsets must be pretty standard to be interchangeable but I cannot find any information on the web - nor, of course, any off-the-shelf device to do what I want. I find it difficult it to believe I am the only one seeking this solution - especially now we are all having to work remotely.
 

AnalogKid

Joined Aug 1, 2013
8,943
The telephone term for that feedback is "sidetone". The main component in an 1880's-to-1970's phone is the "network" - a four-winding transformer-ish thing that interfaces the microphone and speaker to the phone line. It intentionally let the mic winding couple a little energy to the speaker winding so you could tell 1) if the phone was working; and 2) how lound you were talking. It turned your brain into an automatic volume control processor. Very cute. Early cell phones had none, and it was very disconcerting.

To do what you want the full-blown way, you build/buy a small audio power amp circuit, like something based on an LM386. This goes between the computer audio output and the ear-piece of the headphone. There need to be two inputs to this amp, the original audio signal from the computer and the signal coming from the microphone cartridge. But that signal is very low compared to the earpiece signal, so you also need a small mic preamp.

So, a mic preamp, a two-input mixer, and a speaker amp. It sounds like you worked out the system basics on your own, and yes, it really is that complicated if you want good audio fidelity everywhere. Does this sound like something you want to tackle? Parts of it can be done with low-cost modules from ebay.

ak
 
Top